The general goal of my 5-year research plan is to develop new statistical procedures and to improve existing statistical methodology with regard to both the design of environmental health research studies and the analysis and interpretation of data emanating from such studies. Three specific areas will be considered: a) the design and analysis of toxicological experiments involving laboratory animals; b) the design and analysis of epidemiologic studies on human populations; and c) risk assessment. A result of my continuing research involvement in the first two areas over the past few years has been to initiate research interest in the area of risk assessment. Risk assessment analyses will only be as valid as the data on which they are based, and that is why this research plan involves all three pertinent components leading ultimately to such analyses. Research dealing with risk assessment will necessarily involve consideration of theoretical and applied problems in dose-level extrapolation and in species-to-species extrapolation.